Cloudware Operations and Management - Guide to Cloud Computing for Business and Technology Managers - page 385

TABLE 18.3
Key Contractual Eleme nts of an Application SLA
Service-level
parameter metric

• Website response time (e.g., max of 3.5 s per user request)
• Latency of Web server (WS) (e.g., max of 0.2 s per request)
• Latency of DB (e.g., max of 0.5 s per query)

Function • Average latency of WS = (latency of Web server 1 +

latency of Web server 2)/2
• Website response time = average latency of Web server +
latency of database

Measurement directive • DB latency available via http://mgmt server/em/latency

• WS latency available via http://mgmtserver/ws/
instanceno/latency

Service-level objective • Service assurance

• Website latency <1 s when concurrent connection <1000

Penalty • 1000 USD for every minute while the SLO was breached

It is also possible for a customer and the service provider to mutually agree
upon a set of SLAs with different performance and cost structure rather
than a single SLA. The customer has the flexibility to choose any of the
agreed SLAs from the available offerings. At runtime, the customer can
switch between the different SLAs.
Table 18.1 describes the amount of acceptable downtime per year for the
corresponding level of availability.

18.6.1 Quality of Service (QoS)

Quality of Service (QoS) is a well-known concept in other areas. For
example, in networking, QoS is defined in terms of error rate, latency, or
bandwidth and implemented using flow control, resource reservation, or
prioritization.
In classic database system operation, QoS and SLAs are mostly limited to
provide reliable and available data management. Query processing typically
aims at executing each query as fast as possible, but not to guarantee given
response times. However, for database services hosted on a cloud infrastruc-
ture and provided as multitenant service, more advanced QoS concepts are
required. Important criteria or measures are the following:

1. Availability : The availability measure describes the ratio of the total
time the service and the data are accessible during a given time inter-
val and the length of this interval. For example, Amazon EC2 guar-
antees an availability of 99.95% for the service year per region, which
means downtimes in a single region up to 4.5 h per year are accept-
able. Availability can be achieved by introducing redundancies: data